Let me start with a confession: I am no engineering whiz, but I like to know how things work. I studied religion, and I often write about art, which is how I became entranced by astrolabes. Their beauty is mesmerizing, but their efficacy as an instrument leaves me perplexed. Imagine a medieval lass trying to ferret out the secrets of a smartphone, or even a dumb phone. Well, that’s how I feel, and I don’t like it.

So I’ll set myself a task: I, a Brooklynite, am going to find my way around an unfamiliar city—Boston—using an astrolabe. To get started, I ask Sara Schechner about this. She is a historian of science with a special interest in the history of astronomy, and she curates the Collection of Historical Scientific Instruments at Harvard University. “It can’t do that,” she says. “Astrolabes,” she explains to this liberal arts major, “aren’t navigation devices. They’re early computers.”

If they don’t spell out routes from oases to caravanserais or chart courses across the Mediterranean, what do they do? They compute, as Schechner puts it, “the where and when.” Knowing this, any good navigator or mariner can then figure out routes—and much, much more.

Instead of dampening my enthusiasm, Schechner stokes it. No wonder astronomers, astrologers (until recent centuries, often one and the same) and religious leaders from Samarkand to Seville have prized astrolabes. They could compute the next eclipse at any location on any date; know where the planets were or had been; and tell when the sun would rise and set, any day, anywhere.

I leave Schechner’s office determined to master this early—possibly the earliest—portable computer.

Astrolabe laser fabricated from wood with a zenith calibrated to the latitude of Houston,Texas
The author holds her astrolabe, laser fabricated from wood with a zenith calibrated to the latitude of Houston, Texas, above; its back side appears above right.

After two months of reading and studying pictures and diagrams, I am not much better off. Nothing I read seems to stick. Clearly, the only way I am going to pick up even the basics of astrolabe tech is not by head but by hand. So I drive to the science building of Central Connecticut State University, about two hours southwest of Boston. There, in an office crammed with books, papers and The Lord of the Rings fan merchandise, I meet astronomer Kristine Larsen. She hands me a toolkit: a workbook and a few articles with links to PowerPoint presentations. Just more reading? “With these,” she says, smiling, “you can actually make your own astrolabe.” Out of paper. 

That sounds dubiously flimsy, but it turns out that paper astrolabes have a long, proud history. Astronomers sometimes included simplified ones in manuscripts. By the early 1500s, as printing presses became widely available, the likes of 16th-century instrument-maker Georg Hartmann in Nuremberg and 15th-century professor Andreas Stiborius in Vienna realized they could fine-tune one perfect proof and then print a whole series of astrolabes cheaply and accurately. Today anyone can simply download software, which is exactly what I do when I get home.

The first thing Astrolabe Generator (www.astrolabe-project.com) asks of me are my coordinates—latitude and longitude. It next has me choose settings—pretty much like picking apps for my phone. I select “shadow square,” an Islamic app to compute heights of structures using trigonometry. Next is how to figure out the direction of Makkah, a task early astrolabe makers never imagined would be done from across an ocean. Then I also opt for the intriguingly anachronistic “unequal hours,” which divides day and night into 12-hour intervals, regardless of season. This means the length of each hour changes daily. Who wouldn’t need an app to track that? It basically allows you to translate solar time into local time, even compensating for seasons. And it does this with a graceful set of arcs. So beautiful, so clever!

In seconds the software generates diagrams for the parts of my first astrolabe: the circular plate, the back, the rotating rete and the two pointers, the rule and the alidade. The process is so easy I make a few more using the coordinates of friends and relatives. All the backs look exactly the same. And the fronts—or plates—bear the same markings except for one important difference. On all the plates, the center point marks the north celestial pole. If you’ve seen time-lapse photographs of the night sky with revolving stars around a point, that’s it, and the closest star to that point is Polaris, the North Star. 

Where the plates differ is in the position of the second important point, the zenith. This represents the point directly over your head at your specific latitude.

So in a plate calibrated to Houston, Texas, for example, the distance between the zenith and the North Star is greater than in one calibrated to Providence, Rhode Island, which is at a much more northern latitude. (Thinking of it this way helps: The farther north you go, the higher the North Star climbs in the sky until, at the pole, it matches the zenith; Go south, and the North Star falls toward the horizon until, past the equator, it is no longer visible.)

On the plate, both the North Star (at the center) and the zenith are inside a network of lines that map a dome whose apex is the zenith. The arced lines are the equivalent of latitudes and longitudes, but on sky maps they’re called azimuth and almucantar, both reminders of the legacies of Arabic-speaking scientists in both astronomy and astrolabes.

Then there is this mindblower: Opposite of a compass, the astrolabe’s south is at the top of this sky map, and north at its bottom; this makes east left and west right. It takes a while, but I finally get why: As with all historical astrolabes, mine are made for the northern hemisphere, where the point over your head—the zenith—is south of the north celestial pole. So why not just draw the sky map in the lower portion of the instrument and keep north at the top? The answer is because it wouldn’t be as efficient: When measuring the height of a star above the horizon—its altitude—you have to hold the astrolabe vertically, like a pendulum. In this way you can look through the sights on the alidade and line them up with the star to determine the star’s angle from the horizon. Having the sky map on the top makes this a lot smoother, more elegant.  (An astrolabe calibrated in the southern hemisphere would use as its center point the south celestial pole, and then north would be at the top.) 


Then we have the rete. This is what looks like an openwork sculpture. It has two circles: a large one, centered on the north celestial pole, with little pointers coming off it to mark individual stars, and a smaller one, centered on the zenith. This one is all about the Sun, which from our vantage point indeed appears to be circling the Earth. When early Greek astronomers plotted its trajectory across our sky, they used the constellations of the Zodiac as milestones. That is why the rete’s inside circle is divided into 12 zodiac signs, each in turn divided into 30 days. (Because the Sun’s path occasionally passes behind our Moon, Greek astronomers dubbed that path the ecliptic.) 

This same succession of Zodiac signs appears also on the rim of the back of the astrolabe with, beneath it, the calendar we use every day. On Western astrolabes it is usually the Gregorian calendar; on Islamic astrolabes it is usually the hijri calendar. Figuring out the “where and when” begins here, on the back of the astrolabe, where a date can be translated into its astronomical/astrological equivalent.

As I try to absorb all this information, I find myself really wishing I could hold onto something heftier than paper. That’s right: I want a solid astrolabe.  But it has to be calibrated to a us location, and not expensive. So no to replicas on Amazon that won’t really work anyway. No to commissioning a master craftsman. Yes to locating a fabricator. One call leads to another until I find Brandt Graves in Brooklyn. He has never heard of astrolabes, he says, but is game to try to make one. After he loads and formats my files into his computerized laser cutter, we watch it slice and etch a panel of birch plywood. As it zips back and forth, I marvel at the astrolabe makers of old achieving this kind of precision by hand.

Then comes time to pin the components together through the celestial pole. “Gorgeous,” I think to myself. “And, look, the rete swivels.” But I can’t fully absorb that I have just made an interactive sky map until I get it home and start to tackle my  first question: What time does the sun rise and set? I test my new Houston-calibrated astrolabe and try for May 6, 2019, the projected first full day of Ramadan. 

I’m sparing you my litany of false starts—let’s just say that the only way to form good habits is by repetition. Starting at the back of the astrolabe, I find May 6. Once lined up with the date, the alidade points to about Taurus 16¼–17¼ (this means the 24 hours of May 6 begin on the 16th day of Taurus and ends on the 17th day, plus ¼ for each). Turning the astrolabe over, I look on the rete for Taurus 16. I find it, and then I rotate it until Taurus 16 touches the eastern horizon line. (I remember, in the left quadrant). Then nudge it a smidge to account for the ¼. Holding the rete in place, I line up the rule. On the rim is the time. The numbers denote hours, each divided into four-minute gradations: It reads 5:20 a.m. Now this is solar time, of course, and by May, us Daylight Savings Time will be in effect, so I add one hour. Then there is the time zone issue: I have to compensate for the difference between Houston’s longitude, 95 degrees west, and that of the longitude to which its time zone is pegged, 90 degrees west. For every degree I need to add four minutes: a total of 20. Finally, I have to find the “equation of time chart” that compensates for arcane astronomical eccentricities, and it says that I have to add three more minutes for May 6. 

My prediction is … my mental drumroll is deafening … 5:20 a.m. plus one hour, plus 20 minutes, plus three minutes: 6:44 a.m. Without daring to breathe—yes, this is how tense I have gotten—I check online. The sunrise prediction for May 6 in Houston: 6:35 a.m. Nine minutes off. Not terrible, but it won’t get me an apprenticeship in Alexandria. 

So I try sunset. Because the day is well on its way by sundown, I line up Taurus 17 with the western horizon, and I read 6:36 p.m. Now—I’m starting to get the hang of it—I add one hour, 20 minutes and three minutes, which brings me to 7:59 p.m. Online? 8:01! I let out a cheer! 

So do I try to get it even closer? You bet. This time I squeeze the rete so tightly to the plate my fingertips start to turn white. I don’t want a micron of wiggle. I even use a magnifying glass to make sure Taurus 17 crosses the horizon exactly. The rule points this time to 6:38. Siri and Alexa, meet your living ancestor Astrolabe. Spot on!

Feeling empowered, I take on geolocation: Where on the horizon will the Sun be? Remember the sky map’s network of almucantar and azimuth? Well, they’re all separated by a number of degrees—here five. Noting where Taurus 16-ish meets the horizon, I count 14 azimuth, making it 70 degrees east of north. This takes some doing because the wood rete blocks my view in places. And then two things occur to me: First, to get really good at this, I don’t just have to learn how the apps work, I have to get to know the physical peculiarities of my own astrolabe. Second, astrolabe makers would be agog at the possibility of making a rete out of a transparent material like acetate or clear plastic. With paper astrolabes we can do this, and it makes counting the degrees incredibly easier. 

The author guides her stepdaughter, Isa, as she sights along the astrolabe’s alidade to calculate the height of a balcony in an auditorium at Central Connecticut State University.
The author guides her stepdaughter, Isa, as she sights along the astrolabe’s alidade to calculate the height of a balcony in an auditorium at Central Connecticut State University.

But some astrolabe apps require something entirely solid—such as an alidade with raised sights like those on a rifle. Good thing I happen to have one! With these applications I can figure out the height of a building or, say, a mountain. Another app tells time: Here, the trick is to measure the altitude of the Sun without staring into it. On my wood astrolabe, I use pushpins as makeshift sights so that when I turn the astrolabe sideways to the Sun, its shadow looks like a stick with two bumps on it. I then rotate the alidade until the sights align to cast a single shadow, and I look to see where it points along the scale at the outermost ring. It marks degrees from 90 (at the top) to zero. The alidade points to 32. So, the first of February translates into Pisces 1. Now turn the astrolabe over. Find the rete’s Pisces Decan 1 marker. Move it up from the horizon to 30 degrees (sixth almucantar) and a titch beyond. Place the rule there and see where it points … uh, but in which quadrant? Left makes it 9:04 a.m. solar time; right, 3:04 p.m. That’s correct: the astrolabe relies on my knowing a few things, like whether it’s afternoon or morning. 

But what if I was to do this at night? Caravaneers of yore knew the night sky like oases along their routes. As long as it was clear enough to spot one or two of the stars noted on their rete, they were good. They’d sight them with the alidade and figure out “when” it was. Mariners who had planetary charts and tables could also figure out “where” they were, at least in terms of latitude. Me? I’ll blame it partly on living in New York that I haven’t the slightest which twinkles are Altair or Sirius or Deneb. So, yes, I can use 21st-century technology to make an astrolabe, but if I ever hope to truly master one, my own mental databases will need an upgrade.